sign in sign up
<<
Home , Follicular fluid, Ovary, Theca externa, Theca interna

Proc. Nati. Acad. Sci. USA Vol. 85, pp. 1957-1961, March 1988 Medical Sciences Human ovarian cells are a source of (/renin-angiotensin) YUNG S. Do, ANDREW SHERROD, ROGERIO A. LOBO, RICHARD J. PAULSON, TATSUO SHINAGAWA, SHENGWU CHEN, SIRI KJOS, AND WILLA A. HSUEH Departments of Medicine, Pathology, and Obstetrics and Gynecology, Los Angeles County/University of Southern California Medical Center, Los Angeles, CA 90033 Communicated by Irvine H. Page, October 28, 1987

ABSTRACT Human ovarian contains hypothalamic renin system appears to regulate and renin-like activity. In normal women, circulating levels of antidiuretic production and to control dipsogenic prorenin, the biosynthetic precursor of renin (EC 3.4.23.15), behavior (10, 11). An adrenal renin system exists that change in parallel with changes in during the regulates production by the . Therefore, the ovary has been implicated as a (12). source of plasma prorenin. In the present studies, we report Recently, renin-like activity and even larger quantities of a the finding of high concentrations of prorenin in human trypsin-activatable renin have been found in human ovarian ovarian follicular fluid (3000 ng-ml-P'hr- vs. 10-40 ng- follicular fluid (13, 14). Prior to activation this renin cross- ml1X hrXC in normal human plasma) obtained from follicles of reacted with antibody generated against a consisting women prepared for in vitro fertilization. The inactive renin- of the 12 carboxyl-terminal amino acids of the human renin like enzyme was identified as prorenin by its activation char- prosequence, confirming its identity with prorenin (14). acteristics, its molecular weight of 47,000, which is the same as The objective of the present investigation was to localize that for recombinant prorenin, and its cross-reactivity with the site of renin production in the human ovary. human renal renin antibodies. Culture of isolated human theca cells and isolated granulosa cells indicated that prorenin METHODS is secreted by theca cells but not by granulosa cells. Prorenin production by theca cells peaked during the first 10 days of Four women were treated with human menopausal gonado- culture and gradually decreased by 17 days. Active renin levels tropin (Pergonal, 2 amps daily for 5 days; clomiphene, 100 were 10% or less of the prorenin levels. Prorenin was barely mg daily for 5 days) on days 3-9 of their menstrual cycle in detectable in medium from granulosa cells cultured for 24 preparation for in vitro fertilization (IVF). On days 12-16 of days. Immunohistochemical staining of human (n = 5) the cycle after human chorionic administra- with anti-human renin antibody demonstrated the presence of tion, ovarian follicular fluid was aspirated from mature renin primarily in theca cells. These studies suggest that the follicles. Plasma was also obtained from these subjects for theca cell is the source of the large quantities of prorenin in measurement of renin. Active renin and prorenin concentra- human ovarian follicular fluid. tions were measured as described (15). Reversible acid activation of follicular fluid prorenin was tested (15) and The production of the enzyme renin (EC 3.4.23.15) by the trypsin activation of renin (14) was compared to results kidney regulates a cascade of events in the circulation that obtained with acid activation. ultimately generate angiotensin II (AII). This octapeptide The immunoblot methodology has been described in detail vasoconstricts vascular and stimulates aldo- (16). Rabbit antiserum developed against pure human renin sterone biosynthesis, thereby establishing the renin-angio- (16) was used to detect immunoreactive forms of renin. This tensin system as a major determinant of blood pressure (1). antiserum had a 50%6 inhibitory titer of 1:30,000 against 1 x However, extrarenal renin-angiotensin systems may have l0-4 Goldblatt unit of MRC human renal renin and pure other diverse physiologic effects (2). In normal humans, renal renin. Immunohistochemical staining of normal human more than one-half of the circulating renin exists as the kidney using this antibody demonstrated staining localized inactive biosynthetic precursor prorenin (3). Although the to the juxtaglomerular area only. Furthermore, the antibody kidney appears to be a primary source ofthe prorenin as well did not cross-react with trypsin, pepsin, human renal kalli- as active renin in plasma, immunohistochemical and in situ krein, or cathepsin D. hybridization studies have identified potential extrarenal Granulosa and theca cells were isolated and cultured from sites of renin production. These include pituitary, adrenal, premenopausal ovaries obtained in the course of benign testis, , and - (2, 4-8). Human chorion- pelvic surgery by the method of Lobb and Dorrington (17). decidua produces and secretes primarily prorenin (7, 8). Granulosa cells were obtained by gentle scraping ofthe inner Whether active renin or prorenin is the major form of renin surface of a follicle, centrifuged, and plated (105 cells per ml) produced by other extrarenal sources is unknown. However, in Dulbecco's modified Eagle's medium and Ham's F-12 several years after nephrectomy, plasma prorenin levels, but medium (1:1) containing 10o fetal calf serum. The theca not active renin levels, increase in anephric subjects (9). layer was peeled from the stroma, digested, centrifuged, and Thus, in the absence of the kidney, extrarenal sources are plated. Stimulated granulosa cells were also obtained after capable of secreting prorenin. While renal renin production centrifugation of ovarian follicular fluid obtained from IVF may regulate overall vascular homeostasis, regulation at a patients. In addition to active renin and prorenin, 4-andro- paracrine or autocrine level has been assigned to these local stene-3,17-dione and progesterone (18) were measured in the renin-angiotensin systems (2). For example, a pituitary and culture medium. Immunohistochemical studies using the anti-human renal were on normal tissue The publication costs ofthis article were defrayed in part by page charge renin antibody performed ovarian payment. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. §1734 solely to indicate this fact. Abbreviations: All, angiotensin II; IVF, in vitro fertilization.

1957 Downloaded by guest on October 1, 2021 1958 Medical Sciences: Do et al. Proc. Natl. Acad. Sci. USA 85 (1988) obtained from five premenopausal women undergoing ovari- ectomy for ovarian cysts. Ovarian tissue was fixed in 10% formalin and embedded in paraffin. After incubation with rabbit anti-human renal renin antiserum (1:100, 48 hr at 40C), sections were developed with avidin-biotin complex system 44,000 (Vector Laboratories). N. RESULTS 22,000 Z Human ovarian follicular fluid contained nearly 20 times the 18,000 normal plasma levels of active renin and 100 times the plasma levels of prorenin (Table 1). Acid treatment or trypsin treatment of follicular fluid yielded similar values of 1 2 prorenin. Acid activation of follicular fluid prorenin was FIG. 1. Immunoblot analysis of pure active human renal renin reversible. Reexposure to acid renewed enzyme activity. compared to ovarian follicular fluid renin. Lane 1, pure human renal Anti-human renal renin antibody (104 dilution) inhibited 50% renin has a Mr of 44,000 with Mr 22,000 and 18,000 subunits. Lane ofthe enzyme activity of 10-4 Goldblatt unit ofeither acid or 2, only a Mr 47,000 protein in ovarian follicular fluid cross-reacts trypsin-activated follicular fluid renin. with the human renin antibody. Immunoblot analysis of ovarian follicular fluid demon- strated that a Mr 47,000 protein cross-reacted with antisera right ovary from the same patient and also cultured for 24 generated against pure human renal renin (Fig. 1). Thus, days. Overall, prorenin production was markedly higher in follicular fluid renin had a larger molecular weight than pure cultured theca cells from the left ovary compared to the right renal active renin. ovary (Fig. 4). Peak prorenin production occurred at days Supernatant from cultured granulosa cells obtained either 6-8 of culture. The pattern of active renin production was from premenopausal ovarian tissue or from IVF follicular similar. However, active renin only represented -4/10th or fluid contained extremely low quantities of active renin or less of the total renin produced by the theca cells. All prorenin after 1 week of culture (Fig. 2). In contrast, eight cultures contained ::z106 cells per well at 1 week. No de- culture wells of theca cells obtained from follicles of three tectable active renin or prorenin was found in control me- premenopausal ovaries contained 3.2 ± 7 ng-ml-1hr-1 of dium. active renin and 68.3 ± 14.9 ng-ml-1hr-1 of prorenin per The histologic identification of cells in hu- well after 1 week of culture. Differences between granulosa man ovary was established first in hematoxylin and eosin- and theca cell cultures were confirmed by morphological stained sections (Fig. SA). A developing follicle was recog- assessment and measurements of 4-androstene-3,17-dione nized structurally as a cystic space lined by stratified small and progesterone in the medium. When observed under cuboidal cells corresponding to the layer and phase-contrast microscopy, cultures of both theca and gran- containing an eccentrically placed . The layer ofplump ulosa cells appeared homogenous (Fig. 3). Granulosa cells spindle-shaped cells immediately peripheral to the granulosa were polygonal and contained multiple lipid droplets. Theca cell layer was identified as theca interna. The concentric cells were smaller and spindle shaped. The granulosa cells compact layer of fibroblasts and collagen just outside the produced relatively high levels of progesterone and low theca interna was identified as . These compo- levels of 4-androstene-3,17-dione compared to theca cells nents of the developing follicle were easily recognized on (Fig. 2). Thus, the 4-androstene-3,17-dione/progesterone corresponding hematoxylin counterstained sections used in ratio was 6 for the granulosa cell medium compared to 597 the immunohistologic localization of renin. In a follicle con- for the theca cell medium. This is consistent with the known taining a maturing ovum, immunostaining was most promi- relative pattern of production in ovarian granulosa nent in the theca interna with minimal staining occurring in and theca cells (19). the granulosa cells and theca externa (Fig. 5 C and D). This Low levels of renin and prorenin were also found when staining was absorbed out by pure human renal renin (17) and culture medium was sampled every 2-3 days for up to 24 was negative in the presence of preimmune sera (Fig. SB). days from granulosa cells obtained from ovarian tissue of a Staining was generally seen in the cytoplasm of the theca premenopausal patient (Fig. 4). Theca cells were removed cells. Four other ovaries from premenopausal women showed from a 3-cm follicle in the left ovary and a 1-cm follicle in the similar localization of renin. Table 1. Renin activity in ovarian follicular fluid DISCUSSION Renin, ng'ml - 1-hr- This investigation demonstrates that human ovarian theca Ovarian cells contain and produce renin. Both immunohistochemical follicular Human staining of the developing human and culture fluid plasma of isolated human theca and granulosa cells support this Untreated 186 ± 20 5.7 ± 0.9 conclusion. Theca and granulosa cells were identified by pH 3.3 dialysis 3333 + 526 23.0 + 4.3 light microscopy by their location and architecture. Immune pH 3.3 dialysis; neutralize and staining was greatest in the area of the theca interna and was incubate 3 hr at 37°C, pH 7.5 244 ± 40 6.9 + 1.0 minimal in the granulosa layer. pH 3.3 dialysis; neutralize and Measurements of4-androstene-13,17-dione and progester- incubate 3 hr at 37°C, pH 7.5; one in the culture medium confirmed that the isolated cells in pH 3.3 dialysis 3240 ± 216 21.0 + 5.0 culture were either predominantly theca cells or granulosa Trypsin treatment 3570 ± 410 22.5 ± 4.0 cells. Morphological examination of the cells in culture Mean ± SEM of four measurements. Trypsin treatment was indicated that the cultures were, in general, homogenous. performed with trypsin at 50 ,ug/ml for follicular fluid and at 1 The granulosa cell appearance was consistent with previous mg/ml for plasma (13). This was the optimum concentration of descriptions of cultured human granulosa cells (20). Granu- trypsin determined to activate renin in plasma and ovarian follicular losa cell medium contained little prorenin. Theca cells had a fluid. distinctly different appearance compared to granulosa cells Downloaded by guest on October 1, 2021 Medical Sciences: Do et al. Proc. Natl. Acad. Scli. USA 85 (1988) 1959

C 90 B 900 45 E N 0Ia N 0 c E 60 600 c 30 - 0 30 4c1% _0 0 C A C 0 CS. 0 T ...... w 2 30 300 Xl 15 S c 0 Ex 4c U.

4

0 . 0 0 GRANULOSA CELLS THECA CELLS FIG. 2. Renin (A, active; B, prorenin) and steroid (C, 4-androstene-3,17-dione; D, progesterone) production in cultured human ovarian granulosa and theca cells. Mean ± SEM is represented. Theca cells produce relatively large quantities ofprorenin and 4-androstene-3,17-dione, while granulosa cells produce relatively larger quantities of progesterone. (Fig. 2). Prorenin production by human theca cells peaked versible acid activation, a specific characteristic of prorenin after 1 week of culture. Media levels ofprorenin were higher obtained from human kidney and plasma (14) and of recom- in cultures of theca cells obtained from the dominant follicle binant prorenin (21). Exposure to low pH has been demon- compared to theca cells obtained from a smaller follicle. strated to induce a conformational change in human prorenin Large quantities ofinactive prorenin have previously been that exposes the active site without hydrolysis of a peptide demonstrated in ovarian follicular fluid obtained from bond (22, 23). Prorenin then folds back to the inactive state women being prepared for IVF (14). Our studies further with incubation at 370C and neutral pH, which thus reverses confirm that prorenin is the major form of renin in ovarian enzyme activity. Immunoblot analysis indicates that follicu- follicular fluid, contributing t95% of the total renin concen- lar fluid renin has a Mr of 47,000. We report here that the tration. Ovarian follicular fluid prorenin demonstrates re- molecular weight of human follicular fluid prorenin is the same as that found for recombinant prorenin and for human chorionic and amniotic fluid inactive renin (21), all of which are larger than pure active human renal renin, which does not contain the 46-amino acid prosegment (17). Sealey and her group have also demonstrated that plasma prorenin peaks shortly after the peak during the normal menstrual cycle (24). These data strongly suggest may regulate theca cell renin produc- tion, as they appear to regulate renin production (25). Kim et al. (26) have recently shown that follicle- stimulating hormone and increase renin mRNA and renin activity in the rat ovary. Our data suggesting that 4001

- 300

E 0) C 200 C

0 a- 1007

0- 18 20 22 24 Time of culture, days FIG. 4. Prorenin production by human granulosa (L) and theca cells in culture. All cells were taken from the same premenopausal patient during the of her cycle. Theca cells were FIG. 3. Human granulosa (A) and theca (B) cells in culture after obtained from a 3-cm follicle in the left ovary (e) and a 1-cm follicle 1 week. Granulosa cells are polygonal and contain lipid droplets. in the right ovary (A). Prorenin concentrations represent the prore- Theca cells are smaller and spindle shaped. nin secreted into the medium per 24 hr. Downloaded by guest on October 1, 2021 1960 Medical Sciences: Do et al. Proc. Natl. Acad. Sci. USA 85 (1988)

A

D

C

FIG. 5. (A) Hematoxylin and eosin-stained section of human ovary containing a maturing follicle with ovum. (x 7.15.) (B) Hematoxylin counterstained section incubated with preimmune serum. (x 140.) No evidence of positive immunoreactivity is identified. (C) Hematoxylin counterstained section incubated with rabbit polyclonal and anti-renin antiserum (x 140.) Immunoreactivity is localized to the cytoplasm of cells within the theca with no evidence of positivity within the granulosa cell layer. (D) Higher magnification of C. ( x 230.) prorenin production is increased in theca cells from the to follicular fluid prorenin levels in vivo is unknown com- dominant follicle compared to a nondominant follicle is pared to the number in culture contributing to medium consistent with these observations. The dominant follicle levels. However, estimation of the prorenin production rate develops the capacity to sequester follicle-stimulating hor- per day by 106 theca cells from days 4-12 of culture (Fig. 4) mone in the follicular fluid, which then contributes to its suggests that prorenin production would approach 103 ability to sustain follicle estrogen biosynthesis (27). ng ml - l hr-1, which is a similar order of magnitude as the Several possibilities could explain the lower prorenin prorenin concentration in follicular fluid. concentrations in theca cell culture medium compared to A complete renin-angiotensin system appears to exist in that in follicular fluid. First, the follicular fluid was obtained the ovary, since human follicular fluid also contains angio- from women whose ovaries were "hyper" stimulated to tensinogen and All (14, 28). Recently, All and All receptors ovulate, while the cultured theca cells were obtained from have been identified in rat ovarian tissue (29). The All women during the normal menstrual cycle. Second, the receptors were localized to granulosa and theca interna areas follicular fluid prorenin levels likely represent an accumula- of some, but not all, developing follicles and were absent tion of prorenin produced by theca cells over an unknown from granulosa areas of luteinized follicles. In addition, period of time. Third, the number of theca cells contributing angiotensinogen mRNA has been found in rat ovary in Downloaded by guest on October 1, 2021 Medical Sciences: Do et al. Proc. Natl. Acad. Sci. USA 85 (1988) 1961

1/20th to 1/30th the levels found in the liver (30), suggesting 8. Hsueh, W. A., Maccaulay, L. K., Kjos, S., Koss, M., Shaw, that the ovary can also synthesize renin substrate. The finding K., Anderson, P. W., Shinagawa, T., Sherrod, A., Adams, J. of a complete system in the ovary is anal- & Do, Y. S. (1987) Hypertension 10, 368 (abstract). renin-angiotensin 9. Yu, R., Anderton, J., Skinner, S. L. & Best, J. B. (1972) Am. ogous to identification of the complete system in the testis. J. Med. 52, 707-710. Renin and renin mRNA, All, and converting enzyme have 10. Aguilera, G., Hyde, C. L. & Catt, K. J. (1982) been identified in testicular Leydig cells (5, 31, 32). 111, 1045-1050. The role of prorenin production by ovarian theca cells is 11. Scholkens, B. A., Jung, W. & Riescher, W. (1980) Clin. Sci. unknown. Prorenin may be available to provide the rela- 59, 535-565. tively smaller quantities of active renin found in the ovary, 12. Doi, Y., Atarashi, K., Franco-Saenz, R. & Mulrow, P. J. (1984) Hypertension 6, Suppl. 1, 1-124-1-129. which could then initiate the ultimate generation of All. 13. Fernandez, L. A., Tarlatzis, B., Rzasa, P. J., Caride, V. J., Whether prorenin has physiologic effects other than gener- Laufer, N., Negro-Vilar, A. F., DeCherney, A. H. & Naftolin, ation of All is presently unknown. Recently, Pucell et al. F. (1985) Fertil. Steril. 44, 219-225. (33) have demonstrated that All (1 ,AM) stimulates estrogen 14. Glorioso, N., Atlas, S. A., Laragh, J. H., Jewelewicz, R. & production in ovarian slices from pubertal rats. This is Sealey, J. E. (1986) Science 233, 1422-1424. consistent with their finding of All receptors in granulosa 15. Hsueh, W. A., Carlson, E. J. & Dzau, V. (1983) J. Clin. Invest. 71, 506-517. cells (29). In the study of Pucell et al. (33), All did not alter 16. Do, Y. S., Shinagawa, T., Tam, H., Inagami, T. & Hsueh, progesterone secretion, even though it is a major product of W. A. (1987) J. Biol. Chem. 262, 1037-1043. the granulosa cells. Thus, the effect of All appeared specific 17. Lobb, D. K. & Dorrington, J. H. (1987) Fertil. Steril. 48, to estrogen stimulation. 4-Androstene-3,17-dione produced 243-248. by the theca cells is the major precursor of estrogen pro- 18. Lobo, R. A., DiZerega, G. S. & Marrs, R. P. (1985) J. Clin. duced in the theca and granulosa cells. It is possible that the Endocrinol. Metab. 60, 81-87. 19. Ross, G. T. & Schreiber, J. R. (1986) The Ovary in Reproduc- theca cell renin-angiotensin system regulates pro- tive Endocrinology, eds. Yen, S. S. C. & Jaffee, R. (Saunders, duction in an autocrine or paracrine fashion. Whether All Philadelphia), pp. 115-139. contributes to Leydig cell production is un- 20. Furman, A., Rotmensch, S., Dor, J., Venter, A., Mashiach, known. All may have other effects in addition to steroido- S., Vlodavsky, I. & Amsterdam, A. (1986) Fertil. Steril. 46, genesis in the ovary. Its vasoconstricting and angiogenic 514-517. properties (34) may be necessary for follicular growth. 21. Hsueh, W. A., Do, Y. S., Shinagawa, T., Tam, H., Ponte, P. A., Baxter, J. D., Shine, J. & Fritz, L. C. (1986) Hyperten- Thus, several potential physiologic roles for ovarian renin sion 8, Suppl. 2, 78-82. exist. The present knowledge that human ovarian theca cells 22. Hsueh, W. A., Carlson, E. J. & Israel-Hagman, M. (1981) produce renin now emphasizes the direction offuture studies Hypertension 3, Suppl. 1, 1-22-1-29. of regulation of theca cell renin release and effects of renin 23. Derkx, F. H. M., Schalekamp, M. P. A. & Schalekamp, and All on theca and granulosa cell steroidogenesis. These M. A. D. H. (1987) J. Biol. Chem. 262, 2472-2477. investigations will be important to define the role of the 24. Sealey, J. E., Atlas, S. A., Glorioso, N., Manapat, H. & renin-angiotensin system in human Laragh, J. H. (1985) Proc. Natl. Acad. Sci. USA 82, reproduction. 8705-8709. 25. Pandey, K. N. & Inagami, T. (1986) J. Biol. Chem. 261, We thank Anne Santo and Denise Walters for expert secretarial 3934-3938. assistance and Ariel Vijod, Tamara Ching, and Helen Tam for 26. Kim, S. J., Shinjo, M., Tada, M., Usuki, S., Fukamiza, A., technical assistance. This work was supported by National Insti- Miyazaki, H. & Murakami, K. (1987) Biochem. Biophys. Res. tutes of Health Grant AM30254. W.A.H. is a recipient of National Commun., in press. Institutes of Health Career Development Award Grant AM01035. 27. Erikson, G. F. (1987) Endocrinology and Metabolism, eds. Felig, P., Baxter, J. D., Broadus, A. E. & Frohman, L. A. 1. Page, I. H. & Bumpus, F. J. (1974) Angiotensin: Handbook of (McGraw-Hill, New York), pp. 905-951. Experimental Pharmacology (Springer, New York), Vol. 37. 28. Culler, M. D., Tarlatzis, B. C., Lightman, A., Fernandez, 2. Re, R. N. (1984) Arch. Intern. Med. 144, 2037-2041. L. A., DeCherney, A. H., Negro-Vilar, A. & Naftolin, F. 3. Sealey, J. E., Atlas, S. A. & Laragh, J. H. (1980) Endocrinol. (1986) J. Clin. Endocrinol. Metab. 62, 613-615. Rev. 1, 365-291. 29. Husain, A., Bumpus, F. M., DeSilve, P. & Speth, R. C. (1987) 4. Naruse, K., Marakoshi, M., Osamura, R. Y., Naruse, M., Proc. Natl. Acad. Sci. USA 84, 2489-2493. Toma, H., Watanabe, K., Demura, H., Inagami, T. & Shi- 30. Ohkubo, H., Nakayama, K., Tanaka, T. & Nakanishi, S. zume, K. (1985) J. Clin. Endocrinol. Metab. 6, 172-177. (1986) J. Biol. Chem. 261, 319-323. 5. Deschepper, C. F., Mellon, S. H., Cumin, R., Baxter, J. D. & 31. Pandey, K. N., Misono, K. S. & Inagami, T. (1984) Biochem. Ganong, W. F. (1986) Proc. Natl. Acad. Sci. USA 83, 7552- Biophys. Res. Commun. 125, 662. 7556. 32. Pandey, K. N, Maki, M. & Inagami, T. (1984) Biochem. 6. Gross, F., Schaechtelin, G., Ziegler, M. & Berger, M. (1964) Biophys. Res. Commun. 122, 1337-1341. Lancet H, 914-916. 33. Pucell, A. G., Bumpus, F. M. & Husain, A. (1987) J. Biol. 7. Acker, G. M., Galen, F. X., Devaux, C., Foote, S., Papernik, Chem. 262, 7076-7081. E., Pesty, A., Menard, J. & Corvol, P. (1982) J. Clin. Endo- 34. Fernandez, L. A., Twickler, J. & Mead, A. (1985) J. Lab. crinol. Metab. 55, 902-909. Clin. Med. 105, 141-145. Downloaded by guest on October 1, 2021